|
| 1 | +"""MIME-Type Parser |
| 2 | +
|
| 3 | +This module provides basic functions for handling mime-types. It can handle |
| 4 | +matching mime-types against a list of media-ranges. See section 14.1 of the |
| 5 | +HTTP specification [RFC 2616] for a complete explanation. |
| 6 | +
|
| 7 | + http://www.w3.org/Protocols/rfc2616/rfc2616-sec14.html#sec14.1 |
| 8 | +
|
| 9 | +Contents: |
| 10 | + - parse_mime_type(): Parses a mime-type into its component parts. |
| 11 | + - parse_media_range(): Media-ranges are mime-types with wild-cards and a 'q' |
| 12 | + quality parameter. |
| 13 | + - quality(): Determines the quality ('q') of a mime-type when |
| 14 | + compared against a list of media-ranges. |
| 15 | + - quality_parsed(): Just like quality() except the second parameter must be |
| 16 | + pre-parsed. |
| 17 | + - best_match(): Choose the mime-type with the highest quality ('q') |
| 18 | + from a list of candidates. |
| 19 | +""" |
| 20 | +from functools import reduce |
| 21 | + |
| 22 | +__version__ = '0.1.4' |
| 23 | +__author__ = 'Joe Gregorio' |
| 24 | +__email__ = 'joe@bitworking.org' |
| 25 | +__license__ = 'MIT License' |
| 26 | +__credits__ = '' |
| 27 | + |
| 28 | + |
| 29 | +def parse_mime_type(mime_type): |
| 30 | + """Parses a mime-type into its component parts. |
| 31 | +
|
| 32 | + Carves up a mime-type and returns a tuple of the (type, subtype, params) |
| 33 | + where 'params' is a dictionary of all the parameters for the media range. |
| 34 | + For example, the media range 'application/xhtml;q=0.5' would get parsed |
| 35 | + into: |
| 36 | +
|
| 37 | + ('application', 'xhtml', {'q', '0.5'}) |
| 38 | + """ |
| 39 | + parts = mime_type.split(';') |
| 40 | + params = dict([tuple([s.strip() for s in param.split('=', 1)]) |
| 41 | + for param in parts[1:] |
| 42 | + ]) |
| 43 | + full_type = parts[0].strip() |
| 44 | + # Java URLConnection class sends an Accept header that includes a |
| 45 | + # single '*'. Turn it into a legal wildcard. |
| 46 | + if full_type == '*': |
| 47 | + full_type = '*/*' |
| 48 | + (type, subtype) = full_type.split('/') |
| 49 | + |
| 50 | + return (type.strip(), subtype.strip(), params) |
| 51 | + |
| 52 | + |
| 53 | +def parse_media_range(range): |
| 54 | + """Parse a media-range into its component parts. |
| 55 | +
|
| 56 | + Carves up a media range and returns a tuple of the (type, subtype, |
| 57 | + params) where 'params' is a dictionary of all the parameters for the media |
| 58 | + range. For example, the media range 'application/*;q=0.5' would get parsed |
| 59 | + into: |
| 60 | +
|
| 61 | + ('application', '*', {'q', '0.5'}) |
| 62 | +
|
| 63 | + In addition this function also guarantees that there is a value for 'q' |
| 64 | + in the params dictionary, filling it in with a proper default if |
| 65 | + necessary. |
| 66 | + """ |
| 67 | + (type, subtype, params) = parse_mime_type(range) |
| 68 | + if not 'q' in params or not params['q'] or \ |
| 69 | + not float(params['q']) or float(params['q']) > 1\ |
| 70 | + or float(params['q']) < 0: |
| 71 | + params['q'] = '1' |
| 72 | + |
| 73 | + return (type, subtype, params) |
| 74 | + |
| 75 | + |
| 76 | +def fitness_and_quality_parsed(mime_type, parsed_ranges): |
| 77 | + """Find the best match for a mime-type amongst parsed media-ranges. |
| 78 | +
|
| 79 | + Find the best match for a given mime-type against a list of media_ranges |
| 80 | + that have already been parsed by parse_media_range(). Returns a tuple of |
| 81 | + the fitness value and the value of the 'q' quality parameter of the best |
| 82 | + match, or (-1, 0) if no match was found. Just as for quality_parsed(), |
| 83 | + 'parsed_ranges' must be a list of parsed media ranges. |
| 84 | + """ |
| 85 | + best_fitness = -1 |
| 86 | + best_fit_q = 0 |
| 87 | + (target_type, target_subtype, target_params) =\ |
| 88 | + parse_media_range(mime_type) |
| 89 | + for (type, subtype, params) in parsed_ranges: |
| 90 | + type_match = (type == target_type or |
| 91 | + type == '*' or |
| 92 | + target_type == '*') |
| 93 | + subtype_match = (subtype == target_subtype or |
| 94 | + subtype == '*' or |
| 95 | + target_subtype == '*') |
| 96 | + if type_match and subtype_match: |
| 97 | + param_matches = reduce(lambda x, y: x + y, [1 for (key, value) in |
| 98 | + list(target_params.items()) if key != 'q' and |
| 99 | + key in params and value == params[key]], 0) |
| 100 | + fitness = (type == target_type) and 100 or 0 |
| 101 | + fitness += (subtype == target_subtype) and 10 or 0 |
| 102 | + fitness += param_matches |
| 103 | + if fitness > best_fitness: |
| 104 | + best_fitness = fitness |
| 105 | + best_fit_q = params['q'] |
| 106 | + |
| 107 | + return best_fitness, float(best_fit_q) |
| 108 | + |
| 109 | + |
| 110 | +def quality_parsed(mime_type, parsed_ranges): |
| 111 | + """Find the best match for a mime-type amongst parsed media-ranges. |
| 112 | +
|
| 113 | + Find the best match for a given mime-type against a list of media_ranges |
| 114 | + that have already been parsed by parse_media_range(). Returns the 'q' |
| 115 | + quality parameter of the best match, 0 if no match was found. This function |
| 116 | + bahaves the same as quality() except that 'parsed_ranges' must be a list of |
| 117 | + parsed media ranges. """ |
| 118 | + |
| 119 | + return fitness_and_quality_parsed(mime_type, parsed_ranges)[1] |
| 120 | + |
| 121 | + |
| 122 | +def quality(mime_type, ranges): |
| 123 | + """Return the quality ('q') of a mime-type against a list of media-ranges. |
| 124 | +
|
| 125 | + Returns the quality 'q' of a mime-type when compared against the |
| 126 | + media-ranges in ranges. For example: |
| 127 | +
|
| 128 | + >>> quality('text/html','text/*;q=0.3, text/html;q=0.7, |
| 129 | + text/html;level=1, text/html;level=2;q=0.4, */*;q=0.5') |
| 130 | + 0.7 |
| 131 | +
|
| 132 | + """ |
| 133 | + parsed_ranges = [parse_media_range(r) for r in ranges.split(',')] |
| 134 | + |
| 135 | + return quality_parsed(mime_type, parsed_ranges) |
| 136 | + |
| 137 | + |
| 138 | +def best_match(supported, header): |
| 139 | + """Return mime-type with the highest quality ('q') from list of candidates. |
| 140 | +
|
| 141 | + Takes a list of supported mime-types and finds the best match for all the |
| 142 | + media-ranges listed in header. The value of header must be a string that |
| 143 | + conforms to the format of the HTTP Accept: header. The value of 'supported' |
| 144 | + is a list of mime-types. The list of supported mime-types should be sorted |
| 145 | + in order of increasing desirability, in case of a situation where there is |
| 146 | + a tie. |
| 147 | +
|
| 148 | + >>> best_match(['application/xbel+xml', 'text/xml'], |
| 149 | + 'text/*;q=0.5,*/*; q=0.1') |
| 150 | + 'text/xml' |
| 151 | + """ |
| 152 | + split_header = _filter_blank(header.split(',')) |
| 153 | + parsed_header = [parse_media_range(r) for r in split_header] |
| 154 | + weighted_matches = [] |
| 155 | + pos = 0 |
| 156 | + for mime_type in supported: |
| 157 | + weighted_matches.append((fitness_and_quality_parsed(mime_type, |
| 158 | + parsed_header), pos, mime_type)) |
| 159 | + pos += 1 |
| 160 | + weighted_matches.sort() |
| 161 | + |
| 162 | + return weighted_matches[-1][0][1] and weighted_matches[-1][2] or '' |
| 163 | + |
| 164 | + |
| 165 | +def _filter_blank(i): |
| 166 | + for s in i: |
| 167 | + if s.strip(): |
| 168 | + yield s |
0 commit comments